Phase two of the DRR pilot highlighted the importance of agreed data standards. An appropriate solution for producing machine-executable regulation has not yet been identified.
The FCA (Financial Conduct Authority), BOE (Bank of England) and seven regulated firms have jointly published a viability assessment report on the latest digital regulatory reporting (DRR) pilot.
The DRR project was established in 2018 to explore how technology could make it easier for firms to meet their regulatory reporting requirements, potentially allowing them to automatically supply data requested by the regulators, thereby reducing the cost of collection, improving data quality and reducing the burden of data supply on the industry.
Two phases of the DRR pilot have now been completed, involving collaboration between the FCA, BOE and seven banks: Barclays, Credit Suisse, HSBC, Lloyds, Nationwide, Natwest and Santander.
Phase one of the DRR pilot evaluated the feasibility of creating a new regulatory reporting mechanism. It achieved a DLT-based prototype that improved data consistency and quality and increased the efficiency of regulatory reporting.
Phase two
Phase two began in February 2019 to address some of the gaps identified in phase one, understand the economic viability of DRR, and explore how it could apply to different product groups. It also sought to explore possible third party solutions for generating machine-executable regulation and data definitions, and how these fit into the DRR vision.
While the report does not make recommendations as to how DRR should be implemented, it presents a series of implementation scenarios which model its impact in the UK mortgage reporting domain.
Based on the scenarios, it was assessed that the standalone viability of DRR is likely to be low if limited to mortgage reporting. However, when modelling the extension of DRR to other reporting domains, the realisation of benefits was seen as faster.
This assumption would need to be validated as part of the next steps, once the extension domains are agreed. But, one of the general conclusions was that the business case for DRR is strongest when implemented for multiple domains, and if delivered as part of a major regulatory change or new set of reporting rules.
Nevertheless, a phase two analysis of the costs and benefits of DRR for firms and regulators showed the potential for financial benefits, as well as intangible benefits, such as improving regulatory decision making, helping firms deliver better services for their customers, and improving the transparency of regulation.
Data standardisation
Among the key conclusions, phase two found that the lack of standardisation of the definitions and descriptions of data by firms presents a significant barrier to improving firms’ processes and exploiting new analytical techniques.
“Currently some firms use multiple terms or identifiers to describe the same data. The metadata firms hold about their data was not designed to meet the needs of regulatory reporting. Firms often rely on subject matter experts or key individuals to understand the data they have,” the report says.
In the DRR vision, firm data is digitally tagged and identified according to agreed data standards, so that the same data can be identified easily across firms and systems.
Third party solutions for data standards – such as FpML (Financial products Markup Language) and the ISDA CDM (Common Domain Model) – were found to be close to meeting the requirements for derivatives.
ISDA CDM
The CDM was deployed as part of phase two to help understand the feasibility of firms meeting both position-based and transaction-based reporting requirements from the same trade data, and harmonise reporting triggers so firms report the same information at the same time.
Four data points for two separate transactional derivative reports were described using the CDM’s version of digital regulation, opening the potential for a single data model to cover the production of multiple regulatory reports. This suggested the DRR approach could be scalable across reports – a key indicator of economic viability, the report says.
The digital version of the reporting instructions were linked to the CDM, and the CDM to real test files of operational data provided by firms. To the extent that firms implement the CDM, this could “materially reduce the cost” of implementing an approach like DRR for derivatives reporting.
This work also demonstrated how regulators could change reports by changing digital regulation, while seeking to have a single common data model for a product type, and with reports produced off this single model. This is a key benefit of DRR for regulators.
A process for developing a mortgage data standard was also developed, along with a mortgage data model. The model developed was able describe 10 UK mortgage reports, and in combination with digital regulation could be implemented with a fraction of the data point definitions currently contained in the 10 reports.
However, the DRR mortgage model needs to be validated and applied to a live use case.
Machine-executable
A DRR approach would also require the regulator to publish a digital version – i.e. machine-executable – of their regulatory rules, which reference the data standards agreed with firms. The production of such digital rules from the current natural language version would ideally be automated.
Further, the solution would have to work for all firms, be built on open standards and agnostic of technology, capable of being scaled and extended robustly, and be transparent to both firms and regulators.
The research in phase two revealed that there is currently no third party solution available that meets all these requirements. However, the report says, the market is still evolving and solutions may be developed in the future.
The option of developing a purpose-built solution for producing machine-executable regulations was also explored, but further work is needed in subsequent phases to recommend a preferred approach.
Next steps
One of the agreed next steps following phase two is also for the FCA and BOE to commission a joint review on the legal implications of writing machine-executable reporting instructions.
The FCA and BOE have additionally committed to continue working together on common data standards, and to commission a joint review of the some of technical solutions explored as part of the phase two pilot.
Additionally, the work on analysing the business case for DRR is incomplete. More is needed to understand the costs of setting up and running DRR – particularly for regulators – and to get a clearer picture of its benefits.
“For the full benefits of DRR to be realised, a strategic commitment to DRR by firms and regulators may also be required,” the report concludes.
The DRR viability assessment report is available in full here.
The report was published alongside the FCA’s refreshed data strategy and a BOE discussion paper on improving the effectiveness of its data collection, covered here.
